On Mon, Oct 17, 2016 at 09:46:20PM +1100, Matt Flax wrote:
On 17/10/16 21:37, Charles Keepax wrote:
On Mon, Oct 17, 2016 at 09:42:18AM +1100, Matt Flax wrote:
This patch adds support for the wm8581 codec to the wm8580 driver. The wm8581 codec hardware adds a fourth DAC and otherwise is compatible with the wm8580 codec.
of_device_id data is used to allow the driver to select the suitable software variables from the device tree codec selection. The wm8580_driver_data struct is used to store the number of DACs as well as the required stream names for both codecs.
The snd_soc_dai_driver no longer lists the stream names and for playback the channels_max. These variables are set during i2c probe from the of_device_id supplied wm8580_driver_data struct.
With knowledge of the number of DACs in use, the DAC4 controls, widgets and routes are added as required for DAC4.
The device tree documentation for the wm8580 is altered to list the wm8581 codec support, as is the Kconfig file.
Signed-off-by: Matt Flax flatmax@flatmax.org
<snip>
+static const struct snd_kcontrol_new wm8580_dac4_snd_controls[] = { +SOC_DOUBLE_R_EXT_TLV("DAC4 Playback Volume",
WM8580_DIGITAL_ATTENUATION_DACL4,WM8580_DIGITAL_ATTENUATION_DACR4,0, 0xff, 0, snd_soc_get_volsw, wm8580_out_vu, dac_tlv),+SOC_SINGLE("DAC4 Deemphasis Switch", WM8580_DAC_CONTROL3, 3, 1, 0),
+SOC_DOUBLE("DAC4 Invert Switch", WM8580_DAC_CONTROL4, 8, 7, 1, 0),
+SOC_SINGLE("DAC4 Switch", WM8580_DAC_CONTROL5, 3, 1, 1), +};
Ditto here and so on for the next couple.
That would mean replicating the WM8580_DAC_CONTROL[345] definitions where : WM8581_DAC_CONTROL3==WM8580_DAC_CONTROL3 WM8581_DAC_CONTROL4==WM8580_DAC_CONTROL4 WM8581_DAC_CONTROL5==WM8580_DAC_CONTROL5 Is that what you are after ?
Ah apologies that wasn't what I meant if the register already exists then don't change the name. I was referring to wm8580_dac4_snd_controls I would probably name that wm8581_dac4_snd_controls.
- /* Each dac supports stereo input */
- wm8580_dai[0].playback.channels_max = wm8580->drvdata->num_dacs * 2;
- wm8580_dai[0].name = wm8580->drvdata->name_playback;
- wm8580_dai[1].name = wm8580->drvdata->name_capture;
You can't patch the dai_driver struct like this its a static struct that is shared between all instanciations of the driver it is possible someone could put a 8580 and 8581 on the same board.
Oh - That is a good point. Strangely, the cs42xx8 driver has the same problem ! Hmmm ...
Yeah someone should probably look at fixing that at some point, I will try to find some time to have a gander.
I guess you really have two options here, you could specify the maximum in the dai_driver and then apply constraints in the startup callback to limit things to the correct chip dependant number of channels, or just have two copies of the dai_driver struct one for each chip.
I didn't understand your first suggestion, can you give an example of how to do that in the startup callback ? Sounds like a good idea.
If you look in arizona_startup in sound/soc/codecs/arizona.c we specify constraints for the sample rates based off our clocking, there is an equivalent SNDRV_PCM_HW_PARAM_CHANNELS that you should be able to you to specify the supported channels with.
Thanks, Charles